Turned a booking engine into a repeat-trip engine.
Hello packages independent hotels and local experiences into bookable trips. We rebuilt their data layer, trip-search personalization and post-trip lifecycle so more searches became bookings and more guests came back.
Why do travelers search dozens of trips but leave without booking, and what would make them return to us for their next trip instead of starting over on Google?
- Data audit
- Stack buildout
- Lifecycle architecture
- Growth operations
- Custom booking engine
- BigQuery
- Klaviyo
- WhatsApp Business
- Looker Studio
The problem
Hello had a beautiful booking engine and strong top-of-funnel search traffic. But most travelers searched, compared, and left. When they returned months later, they started again as strangers. The same destination was sold to the same person twice through different acquisition channels.
What the funnel audit showed
- Search-to-book conversion was under 3.5% despite high intent traffic.
- The average traveler searched 4.2 times before booking, but each search was treated as a new session.
- Abandoned bookings were followed by a single generic email with no trip context.
- Post-trip communication stopped after the booking confirmation, leaving the next trip to chance.
What we built
- A traveler data layer that joins search behavior, booking history and trip metadata across the booking engine.
- Personalized search results and abandoned booking emails based on party type, destination and trip timing.
- A post-trip lifecycle that triggers the next trip recommendation at the right moment after return.
- A single reporting layer the product and CRM team share, so marketing and product optimize the same funnel.
The result
Search-to-book conversion rose 1.4 points and the 12-month repeat trip rate doubled. More importantly, the booking engine now learns from every search, not just every completed booking.
- PHASE 1Diagnose
Search-to-book funnel, booking engine events, trip metadata and post-booking lifecycle mapped. Found where intent was lost and where identity failed to follow the traveler.
- PHASE 2Build the trip data layer
Unified trip-search, booking, payment and post-trip events in BigQuery. Traveler profile joined across anonymous search, booking and future intent.
- PHASE 3Personalize the funnel
Search recommendations, abandoned booking recovery and post-trip winback based on destination, party type and trip timing, not generic blasts.
- PHASE 4Ongoing
Continuous experiment cadence on search, booking and post-trip programs. Results reconciled against the booking engine and finance.
Search-to-book rate is measured in the booking engine from unique sessions to completed bookings. Repeat trip rate compares 12-month returning booker cohorts before and after the lifecycle launch. Abandoned booking recovery is recovery revenue from triggered flows divided by abandoners reached. Client-approved.
Read our evidence standardWant the same system inside your stack?
See more cases